Tissue Kallikrein 1 and Alternative Complement Pathway Activation in Hematopoietic Stem Cell Transplant-Associated Thrombotic Microangiopathy

Hematopoietic stem cell transplant-associated thrombotic microangiopathy (HSCT-TMA) negatively affects cancer survivorship with high mortality and significant long-term renal morbidity. It is characterized by microangiopathic hemolytic anemia, thrombocytopenia, and endothelial injury that leads to severe renal failure in patients who have undergone HSCT. In HSCT-TMA, dysregulation of innate immunity and inflammation often leads to overactivation of the alternative complement pathway (AP). Proteolysis of complement factor C3 initiates AP activation, and inflammatory proteases outside of the complement system have been reported to cleave C3. Tissue kallikrein 1 (KLK1), a serine protease that is inactivated by kallistatin, is upregulated in inflamed murine kidneys. We hypothesized that increased KLK1 activity arising during inflammation cleaves C3, contributing to AP activation and renal endothelial injury in HSCT-TMA.

To determine if a relationship between complement activation and KLK1 activity exists in HSCT-TMA, we evaluated plasma levels of activated complement factor B (Ba, a marker of AP activation) and inactivated KLK1 in a cohort of pediatric HSCT patients who had plasma samples obtained prior to and after HSCT. In patients who developed HSCT-TMA, Ba levels were significantly higher at diagnosis compared to pre-HSCT (1499 ± 823 ng/ml vs. 831 ± 388 ng/ml, n=10, p<0.05), consistent with AP activation in HSCT-TMA. In patients who did not develop HSCT-TMA, there was no significant difference in Ba levels at a median of 79 days after HSCT compared to pre-HSCT (745 ± 248 ng/ml vs. 694 ± 284 ng/ml, n=24, p=0.31). Semi-quantitative immunoblotting demonstrated significantly less KLK1 in complex with its inhibitor, kallistatin, among HSCT-TMA patients at diagnosis compared to pre-HSCT (32.5 ± 11.1% vs. 53.4 ± 20.8%, n=5, p<0.05), consistent with decreased KLK1 inactivation in HSCT-TMA. In patients who did not develop HSCT-TMA, there was no significant difference in the amount of KLK1 in complex with kallistatin at a median of 90 days post-HSCT compared to pre-HSCT (36.7 ± 6.7% vs. 46.7 ± 11.9%, n=5, p=0.08). Ba levels negatively trended with KLK1 inactivation in HSCT-TMA patients (Pearson R = -0.53, p=0.11, n=5), but not in HSCT patients without TMA (R = 0.05, p=0.91, n=5), suggesting decreased KLK1 inactivation may contribute to an increase in complement activation in HSCT-TMA. To determine if KLK1 may directly activate the AP, we incubated complement factors with or without activated recombinant human KLK1 (rhKLK1a) and assessed for cleavage by immunoblotting. We found that rhKLK1a cleaved C3 completely to activation product C3b, but did not cleave complement factors B or C5, consistent with KLK1 having an initiating role within the AP.

Because immunosuppression and inflammation during the peri-transplant period can trigger HSCT-TMA, we studied the effects of cyclosporin A (CsA, immunosuppressant) and lipopolysaccharide (LPS, inflammatory endotoxin) on AP activation and KLK1 activity in cellular and animal models. Glomerular microvascular endothelial cells challenged for 24 hours with 25 µg/mL CsA + 100 µg/mL LPS generated significantly more Ba and KLK1 in the culture medium compared to control cells (Ba: 70.2 ± 12.8 pg/mL vs. 47.7 ± 16.3 pg/mL, n=6, p<0.01; KLK1: 201.4 ± 67.2 pg/mL [n=6] vs. 31.6 ± 38.9 pg/mL [n=5], p<0.05). C57BL/6J mice challenged with 30 mg/kg CsA + 5 mg/kg LPS for 24 hours had KLK1 that was less inactivated by kallistatin compared to pre-challenge, although this did not reach significance owing to low numbers of mice used (41.2 ± 5.6% vs. 71.0 ± 4.3%, n=2, p=0.15). Histological examination of kidneys illustrated markedly increased glomerular deposition of activated C3 and shedding of cell-bound thrombomodulin (a marker of endothelial cell health), consistent with AP activation and renal endothelial injury.

Altogether, our data suggest a noncanonical mechanism of AP activation in which KLK1 may add to C3-dependent kidney injury in HSCT-TMA. Whether targeting this mechanism improves outcomes in HSCT requires further study.